diff --git a/higher/dataug.py b/higher/dataug.py
index 431bd28..f492f2a 100755
--- a/higher/dataug.py
+++ b/higher/dataug.py
@@ -633,8 +633,8 @@ class Data_augV5(nn.Module): #Optimisation jointe (mag, proba)
                 self._params['prob'].data = self._params['prob']/sum(self._params['prob']) #Contrainte sum(p)=1
 
         if not self._fixed_mag:
-            #self._params['mag'].data = self._params['mag'].data.clamp(min=0.0,max=TF.PARAMETER_MAX) #Bloque une fois au extreme
-            self._params['mag'].data = F.relu(self._params['mag'].data) - F.relu(self._params['mag'].data - TF.PARAMETER_MAX)
+            self._params['mag'].data = self._params['mag'].data.clamp(min=TF.PARAMETER_MIN, max=TF.PARAMETER_MAX) #Bloque une fois au extreme
+            #self._params['mag'].data = F.relu(self._params['mag'].data) - F.relu(self._params['mag'].data - TF.PARAMETER_MAX)
 
     def loss_weight(self):
         if len(self._samples)==0 : return 1 #Pas d'echantillon = pas de ponderation
@@ -663,8 +663,9 @@ class Data_augV5(nn.Module): #Optimisation jointe (mag, proba)
             return torch.tensor(0)
         else:
             #return reg_factor * F.l1_loss(self._params['mag'][self._reg_mask], target=self._reg_tgt, reduction='mean') 
-            params = self._params['mag'] if self._params['mag'].shape==torch.Size([]) else self._params['mag'][self._reg_mask]
-            return reg_factor * F.mse_loss(params, target=self._reg_tgt.to(params.device), reduction='mean')
+            mags = self._params['mag'] if self._params['mag'].shape==torch.Size([]) else self._params['mag'][self._reg_mask]
+            max_mag_reg = reg_factor * F.mse_loss(mags, target=self._reg_tgt.to(mags.device), reduction='mean')
+            return max_mag_reg
 
     def train(self, mode=None):
         if mode is None :
diff --git a/higher/train_utils.py b/higher/train_utils.py
index c51668d..d22fb31 100755
--- a/higher/train_utils.py
+++ b/higher/train_utils.py
@@ -728,7 +728,7 @@ def run_dist_dataugV2(model, opt_param, epochs=1, inner_it=0, dataug_epoch_start
             if(high_grad_track and i>0 and i%inner_it==0): #Perform Meta step
                 #print("meta")
 
-                val_loss = compute_vaLoss(model=fmodel, dl_it=dl_val_it, dl=dl_val) + fmodel['data_aug'].reg_loss()          
+                val_loss = compute_vaLoss(model=fmodel, dl_it=dl_val_it, dl=dl_val) #+ fmodel['data_aug'].reg_loss()          
                 #print_graph(val_loss)
 
                 #t = time.process_time()
diff --git a/higher/transformations.py b/higher/transformations.py
index d9b1306..0410a78 100755
--- a/higher/transformations.py
+++ b/higher/transformations.py
@@ -145,6 +145,7 @@ def zero_stack(tensor, zero_pos):
     
 #https://github.com/tensorflow/models/blob/fc2056bce6ab17eabdc139061fef8f4f2ee763ec/research/autoaugment/augmentation_transforms.py#L137
 PARAMETER_MAX = 1  # What is the max 'level' a transform could be predicted
+PARAMETER_MIN = 0.1
 def float_parameter(level, maxval):
   """Helper function to scale `val` between 0 and maxval .
   Args: